The cultivated olive tree (Olea europaea L. subsp. europaea var. europaea) is among the most iconic and important crops native to the Mediterranean Basin. By the fifth millennium BCE, cultivation of olives was widespread in the region (Zohary et al., 2012) and had altered the agricultural landscape of the southern Levant (an area covering modern-day Syria, Lebanon, Israel, the Palestinian Authority, and western Jordan). Archaeological evidence suggests that by the Early Bronze Age (4^th-3^rd millennium BCE) trade in table olives and olive oil was extensive (Zohary et al., 2012; Langgut et al., 2019), and Late Bronze Age shipwrecks and texts (such as the 14^th century Ugaritic tablets), show the international character of olive trade by the late second millennium BCE. The olive tree is mentioned numerous times in the Hebrew and Christian bibles, as well as the Quran, demonstrating the importance of the tree to the cultural heritage of people in the southern Levant (Kaniewski et al., 2012). Olives and olive cultivation are mentioned extensively in ancient Roman agricultural texts, including those by Cato (De Agricultura, second century BCE) and Columella (De re rustica, first century CE), as well as in the Hebrew Mishna and Talmud (third century CE). Beyond a daily provision, the biblical text suggests ancient Israelites used olive oil in the ordination of high priests and kings (Exodus 30:33) and as fuel for lamps (Exodus 27:20, Leviticus 24:2, see also Welch, 2022). The olive also symbolized peace and prosperity in the bible (Genesis 8:11, Deuteronomy 8:8), and victory and wisdom in ancient Greece (Therios, 2009).

Several important agricultural crops were first domesticated and cultivated in the eastern Mediterranean, including wheat, barley, lentil, chickpea, pea, flax, and olive (Zohary et al., 2012). The cultivation and domestication of annual grains such as wheat and barley preceded those of fruit trees, which required a different knowledge of vegetative propagation (Zohary and Spiegel-Roy, 1975; Spiegel-Roy, 1986; Larson et al., 2014; Abbo et al., 2015). Crop wild relative (CWR) populations of many species are scattered around the southern Levant, including naturally growing olive trees. These trees, O. europaea subsp. europaea var. sylvestris (Mill) Lehr, are thought to be the ancestor of cultivated olives (Zohary et al., 2012) and are considered ‘genetic founder stocks’ for cultivated olives (Lev-Yadun et al., 2000). Olive landraces, which presumably encompass a higher level of genetic diversity than their modern decedents (cf. Zhang et al., 2017), might thus represent a genetic stock closer to the ancestor gene pool. However, the long history of olive cultivation and likely hybridizations between wild and cultivated variants raises the question whether genuine populations of var. sylvestris still exist (Barazani et al., 2016).

It is generally accepted that the cultivation of olive trees started through selection from natural populations of wild O. europaea subsp. europaea var. sylvestris (Mill) Lehr (Zohary and Spiegel-Roy, 1975; Terral et al., 2004; Kaniewski et al., 2012). It is still unclear whether the domestication of crops was initiated by the conscious selection of desirable traits (e.g., Abbo et al., 2012; Spengler, 2020). Nevertheless, it is reasonable to assume that the cultivation of olives started through the propagation of ‘better’ phenotypes, e.g., trees with larger fruits, high oil content, increased yield, etc. (Zohary and Spiegel-Roy, 1975; Zohary et al., 2012). Additionally, it is broadly accepted that the cultivation of fruit crops in general—and olive specifically—was strongly associated with the invention of grafting (Zohary et al., 2012). However, the ability for vegetative cultivation of olives by removing and transplanting truncheons suggests that olive cultivation might have begun independently from the invention of grafting (Foxhall, 2007).

Several recent publications have summarized the domestication process of the olive tree (Kaniewski et al., 2012; Besnard et al., 2013b; Diez et al., 2015; Besnard and Rubio de Casas, 2016; Díez and Gaut, 2016; Besnard et al., 2018; Langgut et al., 2019). Many aspects of domestication have been debated, including where domestication first occurred, and whether there were multiple domestication events. Besnard et al. (2013b) suggested that the domestication of olive trees started in the northern Levant and was followed by secondary diversification around the Mediterranean Basin. More recently, an alternative scenario describing two independent centers of domestication in the eastern and central parts of the Mediterranean area has been postulated (Diez et al., 2015). The contradicting theories of single vs. multiple independent domestication centers of olive have reinvigorated the debate regarding the origins and processes of olive domestication and cultivation (Besnard and Rubio de Casas, 2016; Díez and Gaut, 2016). A third hypothesis recently put forth by Langgut et al. (2019), based on the palynological record, suggests that the southern Levant was the center of initial cultivation from which other domestication centers evolved. Based on this hypothesis, this review summarizes the history of olive cultivation and usage in southern Levant, from the Chalcolithic through the Bronze and Iron Ages, and emphasizes the potential and importance of combining different scientific disciplines when studying crop domestication in general, and olives specifically. It also summarizes the southern Levantine genetic evidence emphasizing the value of local plant genetic resources, including both the wild olive relative and local landraces.

Naturally growing olive trees show remarkable phenotypic differences from cultivated varieties, possessing significantly smaller fruits (presumably with lower oil content), high variation in fruit morphology (e.g., Galili et al., 2021; Terral et al., 2021), and possessing a bushy nature often with a long juvenile stage ( Figure 1A ). However, the long co-existence of cultivated olive trees alongside their wild relatives, which provided opportunities for hybridization between the two, problematizes whether existing wild populations represent genuine examples of var. sylvestris, or are merely feral populations. Consequently, the genetic structure of modern populations of non-cultivated olive trees can be strongly influenced by cultivation in adjacent olive groves (Barazani et al., 2016).

Using molecular markers, several studies investigated the genetic relationships between cultivated olives and the supposedly wild var. sylvestris in an attempt to better understand the olive tree domestication process (Baldoni et al., 2006; Kaniewski et al., 2012; Besnard et al., 2013b; Diez et al., 2015). In addition, new generation sequencing and molecular marker techniques offer novel means to identify populations of var. sylvestris trees and differentiate between wild and feral populations (Besnard et al., 2001; Breton et al., 2006; Besnard et al., 2013a; Besnard et al., 2013b; Diez et al., 2015; Unver et al., 2017; Gros-Balthazard et al., 2019). Moreover, the assembled genomes of var. sylvestris (Unver et al., 2017) and old cultivated trees of the Farga cultivar (Cruz et al., 2016), as well as several transcriptome studies (Munoz-Mérida et al., 2013; Carmona et al., 2015; Iaria et al., 2016; Sarah et al., 2017), and genotyping by single nucleotide polymorphism (SNPs) (Mariotti et al., 2020) have been recently used as the basis for a wide transcriptome screen of var. sylvestris and cultivated accessions, revealing the genomic consequences of olive domestication (Unver et al., 2017).

In the southeast Mediterranean, the hypothetical southern distribution of the wild var. sylvestris includes the Galilee and the Carmel mountain range (Zohary and Spiegel-Roy, 1975; Zohary et al., 2012). More recently, a species distribution model using temperature and the locations of wild-growing olive groves as defining variables was used by Besnard et al. (2013b) to infer the present distribution of wild var. sylvestris. Notably, the model discarded locations of nearby cultivated groves, thus excluding the possibility of feral populations. The resulting map (Besnard et al., 2013b) inferred suitable habitats in the Galilee, Carmel, Samaria, and Judean mountain ranges. However, as these geographical regions are typical of traditional olive cultivation (see Barazani et al., 2014; Barazani et al., 2017), the resulting map differs from the hypothetical distribution range of wild olives in the southeast Mediterranean suggested by Zohary et al. (2012) ( Figure 2 ).

In a previous study, using genetic diversity parameters derived from 15 SSR (simple sequence repeats markers, microsatellites), Bayesian clustering with the admixture model, and Rousset’s genetic clustering, we differentiated genetically between feral, neglected cultivated olive trees in abandoned groves and populations of what look like wild var. sylvestris (Barazani et al., 2016). In support of the distribution range of wild olives in the region suggested by Zohary et al. (2012), naturally growing trees sampled in the Judean mountains were identified as feral trees (Barazani et al., 2016), as well as in the Negev desert ( Figure 1B ). However, populations from the Galilee and Carmel showed distinct genetic differentiation from feral trees and the main cultivated landraces (below), indicating that wild populations of var. sylvestris still exist in these regions ( Figures 1 , 2 ).

Traditional olive groves in the southern Levant are mainly concentrated in the coastal, inland, and mountainous areas of the southern Mediterranean districts of the region (i.e., modern Lebanon, Israel, the Palestinian Authority, and Jordan), areas with more than 350 mm of rainfall/year (see the distribution map of olives in Israel in Figure 2 ). However, olive exploitation since the Bronze Age (above) expanded the cultivation zone of olives to more semi-arid and arid zones of the eastern Mediterranean (<300 mm rainfall/year), as exhibited by ancient living olive trees in the Negev desert, relicts of Byzantine agricultural system (Tepper et al., 2022) ( Figure 2 ). Thus, the range of environmental conditions in the region (Goldreich, 2003) suggests that human activity throughout the long history of olive cultivation may have selected genotypes exhibiting adaptive tolerance to climatic stress conditions, especially water stress (Barzilai et al., 2021). In the era of drastic global climate change, such clones/cultivars could determine the future of olive production.

Estimations for the number of olive cultivars vary widely, from several hundred (Olea database, http://www.oleadb.it/) to around 2,600 (Rugini and Lavee, 1992). Olive cultivars are distinguished by several morphological characteristics of the fruits and leaves, tree shape, oil characteristics, and their utilization (e.g., use for oil or table olives). In recent years, the use of molecular marker techniques, especially SSRs and SNPs, proved to be reliable for cultivar differentiation and identification (Rallo et al., 2000; Sefc et al., 2000; Cipriani et al., 2002; Essadki et al., 2006; Besnard et al., 2011). Thus, the employment of molecular marker techniques proved efficient in characterizing olive germplasms (Khadari et al., 2003; Loumou and Giourga, 2003; Belaj et al., 2003a; Belaj et al., 2003b; La Mantia et al., 2005). These studies included molecular identification of traditional local southeast Mediterranean cultivars while also revealing the genetic relationships among cultivars originating from around the Mediterranean Basin (Owen et al., 2005; Haouane et al., 2011; Díez et al., 2016; and others).

Using a multi-locus lineage (MLL) analysis of 15 SSR markers, we previously reported on a genetic survey of ancient living olive trees in Israel and the Palestinian Authority (Barazani et al., 2014). All trees included in the survey were selected to represent old trees (assumed by trunk perimeter >2 m) growing in traditional olive groves. The MLL analysis considered a mutational threshold, thus excluding genetic differences due to somatic mutations, which enabled the identification of clonal identity. The findings indicated that the majority of living old olive trees (90%) in the southeast Mediterranean region belong to a single MLL (MLL1), one associated with the most common and widely dispersed east Mediterranean cultivar, Souri (Barazani et al., 2014). The Souri cultivar, typical to the southern Levant, occupies most of the traditional rain-fed olive groves. Thus, it is highly adaptable to varying climatic and semi-arid conditions, shallow and stony marginal soils (Ben-Ari et al., 2014), and occasional droughts.

Twenty-seven other local olive cultivars were described in the first half of the 20^th century by Goor (1948). These are divided in current terminology into five main cultivars: Nabali Baladi, Nabali Muhassan, Mailsi, Souri, and Souri Rumi (Barazani et al., 2008). Local olive growers consider the latter cultivar to be a remnant of olive trees from the Roman period. Supporting the existence of specific unique clones/cultivars, the MLL analysis identified the existence of several other clonal groups among ancient living trees, as well as a second large MLL group (MLL7) that was pronounced as a rootstock of old grafted trees (Barazani et al., 2014). Testing the potential contribution of MLL7 as a rootstock, a linear regression analysis revealed that the spread of MLL7 in the southeast Mediterranean decreases with increasing aridity (Barazani et al., 2017). Thus, the distribution of MLL1 grafted onto MLL7_rootstock in old olive trees was more pronounced in high elevations mesic regions (Galilee and Carmel), whereas the distribution of the Souri cultivar (MLL1) as non-grafted old trees increased in regions outside the natural Mediterranean olive cultivation zone. In previous studies (Tugendhaft et al., 2016; Barzilai et al., 2021), we tested MLL1’s response to drought compared to Barnea cultivar trees of the same age and the Spanish drought-resistant Picual cultivar (Shaheen et al., 2011). Monitoring several physiological traits including stomatal conductance, net photosynthesis, leaf water potential and stem growth (Tugendhaft et al., 2016) and soil volumetric water content, stem water potential, and gas exchange (Barzilai et al., 2021), the results of the two studies pointed to the higher drought tolerance of the Souri cultivar. Thus, the selection of the Souri cultivar supports the hypothesis that the selection of the Souri cultivar enabled the expansion of the olive cultivation zone in the southeast Mediterranean into more arid habitats.

The grafting of olives is thought to increase the survival of the propagated trees (Foxhall, 2007), thus improving the propagation success of cultivars that do not root easily, such as the south Levantine Souri (MLL1) cultivar. However, the rooting capability of MLL1 was not significantly different in comparison to MLL7 ( Figure 3 ), suggesting that MLL7 does not facilitate propagation. The use of certain wild-growing olive trees as a source of rootstock that increases tree vigor was previously reported in Turkey (Zohary et al., 2012). Rootstock genotype has been shown to influence olive growth and tolerance to Fe deficiency in calcareous soils (Alcantara et al., 2003). In addition, the results of an Akaike information criterion (AIC) model selection procedure also indicate that grafting of the common Souri cultivar (MLL1) on MLL7_rootstock positively improved oil quality under certain environmental conditions (Barazani et al., 2017). However, whether grafting increased tree vigor in the southern Levant, or provided an easy means for propagation, remains unclear.

As an additional line of investigation, we report here on the genetic identity of 124 olive trees growing in traditional groves in the southern Mediterranean. The clonal identity of the trees was tested in reference to their cultivar name, identified by the growers as Souri, Souri Rumi, Nabali Baladi, Nabali Muhassan, and Mailsi ( Figure 4A ). The results of genetic clonal identification (MLL analysis) revealed that all trees that the growers identified as Souri, Souri Rumi and Nabali Baladi, as well as three trees that the growers identified as Malisi, were clustered together in the MLL1 genetic group ( Figure 4B ), thus supporting synonymous terminology. A second sizeable genetic group included trees belonging to the Nabali Muhassan cultivar ( Figure 4B ). It is important to note that during sampling, the growers and owners of the olive groves pointed to trees that were propagated from the same cultivar but possess different phenotypic traits and thus were named differently. In addition, four trees identified by the growers as being Malisi were separated into single occurrence MLLs, different from the main three local genetic groups, i.e., MLL1 namely Souri, MLL7, and Nabali Mohassan, as well as the reference cultivars Barnea, Arbequina, Coratina, Koroneiki and Picual ( Figure 4B ). Thus, the genetic survey of the local olive germplasm ( Figure 4 ; Barazani et al., 2014) identified unknown clones/cultivars with presumed agronomical potential that has yet to be studied.

Overall, the tangible evidence summarized above indicates that the earliest indication of olive cultivation was found in the southern Levant. As this part of western Eurasia is generally considered one of the domestication hotspots for many crops (e.g., emmer wheat, barley, garlic, lettuce, among others) (Zohary et al., 2012), studies that aimed to decipher olive domestication, (e.g., Lumaret et al., 2004; Breton et al., 2008; Besnard et al., 2013b; Diez et al., 2015) included local cultivars, and several naturally growing olive populations from the southeast Mediterranean. These included local landraces (Souri and Nabali) but, in some cases, also modern local southeast Mediterranean varieties such as Barnea, Merhavia, and Ma’alot (see Besnard et al., 2013a and Diez et al., 2015). In addition, we previously showed that adjacent olive cultivation could strongly influence the genetic structure of modern populations of non-cultivated olive trees (Barazani et al., 2016). Thus, such studies may have wrongly used biological material from feral populations. We suggest that our genetic survey of naturally growing olive populations (Barazani et al., 2016), the unique local olive clones identified by our molecular approach (above), may represent previously unknown ancient genetic stock of early cultivated olives and thus might be utilized to add to the debates over olive domestication. More importantly, underutilized clones selected to withstand harsh local conditions can be reintroduced as better-adapted crops fitted to the changing environment or used as a source for desirable new traits in breeding programs.

OB and ZD wrote the manuscript with contribution from AD. All authors edited and approved the final manuscript.